The requirements for the side channel evaluation of automotive products are emphasized and it is discussed how the proposed methodology meets these requirements. An evaluation methodology suitable for the application in the automotive domain must meet four major requirements. First, it must be possible to fully automate the methodology by integrating it into the design flow. Because any design change can introduce a new side channel vulnerability, the evaluation must be repeated frequently throughout the entire product development cycle. Second, the computational cost of the evaluation must be kept low. Third, the methodology must account for an economically motivated adversary by applying a bounded attacker model. For side channel attacks in the automotive domain, only the measurement effort is relevant because of the extended lifetime of automotive electronics and the constant increase of the computational power due to Moore's law. The computational cost of the attack should be therefore neglected. Finally, the methodology has to include a metric able to express the economic implications of securing a given product using different side channel countermeasures. In contrast to the state of the art, the proposed side channel evaluation methodology allows to identify effective and cost-efficient side channel countermeasures for the product under evaluation. The key to identifying cost-efficient countermeasures is a side channel evaluation in the context of a specific product. The proposed methodology accomplishes this by evaluating the product's side channel resistance in a bounded attacker model that is derived from the model of a given product. The side channel leakage is acquired using a standardized FPGA-based leakage emulation platform which allows to start the evaluation in early design stage. The proposed methodology applies digital signal processing techniques to reduce the computational workload and ensure a prompt availability of the evaluation results. Prior to computing the statistical distinguisher for the subkey hypotheses, the side channel leakage measurements are compressed by transforming the recorded leakage into the frequency domain, computing the power spectral density of the frequency domain representation, and selecting the frequency components which actually contain key-dependent side channel information. The methodology presented herein applies a single statistical distinguisher - namely, the Pearson's sample correlation coefficient - to discriminate the subkey candidates. In contrast to the existing proposals, specifying the distinguisher to be used allows to fully automate the evaluation. Introducing a cost-benefit metric the economical dimension to the side channel evaluation is added.
Side channel evaluation methodology for the automotive domain
Verfahren zur Bewertung von Seitenkanälen im Bereich des Automobils
2014
163 Seiten, Bilder, Tabellen, 243 Quellen
Hochschulschrift
Englisch
Design-Fluss , digitale Signalverarbeitung , Emulation , Evaluation , Frequenzbereich , Gefährdungsanalyse , Hacker (Computer-Kriminalität) , Kfz-Elektronik , Korrelationskoeffizient , Kraftfahrzeug , Kryptologie , Messen , Produktentwicklung , Rechenkomplexität , Schlüssel (Verschlüsselung) , Spektraldichte , statistische Auswertung
Side channel evaluation methodology for the automotive domain
TIBKAT | 2014
|N Channel High-Side Switch - The Solution to Automotive Multiplex
Kraftfahrwesen | 1985
|N Channel High-Side Switch-The Solution to Automotive Multiplex
SAE Technical Papers | 1985
|